Test cases are primary thing that are used for testing software work product. Can we have conscious, unconscious and semi-conscious test cases? Any test case that can take care of >95% the functionality or statements can be called conscious test case. Test case that have code coverage between 70% - 95% can be called as semi-conscious test cases. Test cases with code coverage less than
Hi, Syed
In this context, IMHO, I think that "conscious" is not a good word for an appropriate terminology. "Conscious" reminds some reflexive and smart behaviour - related to intelligence. On the other hand, there is a variety of computational intelligent mechanisms applied to software testing and software engineering. For instance http://cds.cern.ch/record/1701367 and http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=4601522&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D4601522
I agree that the words "conscious," etc., are not good descriptors of what is essentially functional coverage. Test cases, by themselves, cannot be deemed "conscious" by any reasonable interpretation of the term. However, the development of the test cases by the testers can (see below).
There are several aspects to this that bear thinking about:
(a) the functionality to be covered is itself a proposition, not a proof, we might expect the functionality to be there, but it's not;
(b) error masking might delude us into think coverage occurred when it didn't
(c) we cannot legitimately state that test cases cover 95% of functionality unless we qualify that functionality as discrete (ie., only one function) since we can combine sets of functionality in an infinite number of sequences;
(d) while we might be able to claim to cover all expected functionality, it is not possible to cover unexpected functionality (things the system does that it shouldn't do).
So "coverage" is really a relative measure--more coverage is generally better, but we can't assert an absolute coverage.
There are, however, conscious and unconscious actions that the testers can and do take. In fact, I think that the "unconscious" (more correctly: "below conscious")cognitive processes that tester use might well be some of the more powerful and more effective.
Rather than go into this in depth in this narrow column, I attach an article from Communications of the ACM published in January 2005 that, um, covers this issue. Consciously too..
- Badges
- Science topic
- Similar topics
- Geoscience
- Cartography
- Mapping